Gas turbines, once an obscure technology, are now enjoying a bright spotlight driven by AI-related demand. To meet the growing power needs of the AI era, global capital—particularly in the North American market—is focusing on traditional power generation equipment such as gas turbines, gas internal combustion engines, and diesel generators.

Stimulated by surging demand, related A-share stocks are seeing both order volumes and share prices rise sharply. Yingliu Co., Ltd., a supplier of turbine blades for gas turbines, has seen its stock price increase by more than 50% since the start of the year. Meanwhile, Dongfang Electric Corporation, a direct manufacturer of gas turbines, has seen its stock rise over 30% this year.

Institutional analysis suggests that as global data centers continue to expand, U.S. AI data centers alone will require an additional 31 GW of electricity over the next five years. Gas turbines, with advantages such as quick startup, ease of deployment, stable power generation, energy efficiency, environmental friendliness, low maintenance, and cost-effectiveness, may become the optimal power supply solution for AI data centers.

Orders and stock prices are climbing together. Gas turbines are categorized by power capacity into heavy-duty and light-duty types. Heavy-duty gas turbines are mainly used in urban power grids with fixed generating units, while light-duty gas turbines are applied in industrial power generation, marine power, oil and gas transportation, distributed generation, and military applications.

Spurred by electricity demand from ongoing AI data center construction, the gas turbine industry is experiencing a boom in both orders and stock performance. On the evening of February 1, Jereh Group announced that its wholly-owned subsidiary, GenSystems Power Solutions LLC, signed a sales contract with a U.S. customer to supply gas turbine generator sets worth $181.5 million. Since making a breakthrough in the North American market in November 2025, Jereh has secured four consecutive orders in the region, with a total contract value of approximately RMB 3.4 billion.

Boosted by these large orders, Jereh's stock price has performed strongly. Throughout 2025, the company's shares surged 96%, and since the start of this year, the stock has risen another 38.94% in less than two months. Similarly, Yingliu Co., Ltd., which supplies turbine blades for gas turbines, has seen its stock price increase more than 50% this year, while Dongfang Electric Corporation, a direct manufacturer of gas turbines, has risen over 30%.

The sector's prosperity is likely to continue. Demand for gas turbines in the global market is expected to keep growing. Zhongtai Securities predicts that U.S. AI data centers will require an additional 31 GW of electricity over the next five years. Global data center electricity consumption is projected to increase from 415 TWh in 2024 to 945 TWh in 2030, with the U.S. accounting for 45% of this demand. Gas turbines, with their quick startup, ease of deployment, stable output, energy efficiency, low maintenance, and cost advantages, are considered the best choice for powering AI data centers.

Soochow Securities also pointed out that the power shortage in North America stems from the conflict between non-linear growth in AI power demand and aging grid infrastructure. On the demand side, the surge in U.S. AI data center projects has led to rapid increases in electricity needs. On the supply side, while total supply is expected to meet short-term demand in 2025, the U.S. faces challenges in the long term, including declining stable supply and regional power shortages. Due to supply-demand imbalances and rising AI capital expenditure, NERC forecasts an annual peak shortage of over 20 GW in the U.S. from 2027 to 2030, with Texas, the Mid-Atlantic, the Midwest, and California facing significant risks. The DOE predicts an annual peak shortage of 20–40 GW by 2030.

Soochow Securities stated: "Considering cost, construction time, and environmental factors, gas turbines are currently the optimal solution for self-built power at AI data centers, with gas internal combustion engines, SOFCs, and diesel generators serving as effective supplements."

In combined-cycle operation, gas turbines can achieve power generation efficiency of over 60%, with the lowest cost per kilowatt-hour. Current gas turbine installation rates are accelerating, and global new installed capacity in 2025 is expected to approach the peak of the last cycle. Leading manufacturers such as GE Vernova, Siemens, and Mitsubishi Heavy Industries have order backlogs extending to 2029.

Additionally, according to a Gas Turbine World report, global gas turbine sales in megawatt terms are projected to increase from 58 GW in 2024 to 71 GW in 2025, with the number of units ordered rising from 471 in 2024 to 964 in 2025.